Xiangyu Zhu

The safe production of natural gas in reservoir highly depends upon sealing capacity of the conventional and hydrate-bearing caps, with breakthrough pressure serving as the primary evaluation indicator. This study first systematically summarizes the basic principles, procedures, and applicable conditions of five traditional and auxiliary methods for testing conventional caps, namely mercury intrusion porosimetry, the pulse method, the stepwise pressurization method, the continuous injection method, and the displacement method. Building on this foundational understanding, the study then delves into an in-depth analysis of the limitations inherent to these methods when they are applied to hydrate-bearing caps, as well as potential strategies for their improvement. Upon thorough comparative analysis, this study highlights that the stepwise pressurization method, continuous pressurization method, and the pulse method are the potential techniques for determining the breakthrough pressure measurement of hydrate-bearing caps, provided that improvements are made in specimen preparation, temperature-pressure and stress state control. The suitability of these methods depends upon factors such as hydrate saturation and experimental conditions. Finally, this review summarizes current research progress and key challenges in this field. It is proposed that the development of full-process in-situ pressure-core testing systems, the establishment of unified criteria for breakthrough and loading path, and studies on cyclic breakthrough self-healing behavior are crucial steps for transitioning this research from laboratory settings to practical engineering applications.

SUMMARY: Elastic rock physics models are widely used to estimate the saturation of hydrate in isotropic sediments. However, for isotropic media, the influence of heterogeneously distributed hydrate on the P- and S-wave velocities remains unclear, leading to uncertainties in hydrate saturation estimates. To address this issue, in this work we proposed a double-solid-matrix model for predicting the velocities of sediments hosting heterogeneously distributed hydrates. A comparison of simulated velocities of our model and two rock physics schemes designed for homogeneous distributed hydrate (i.e. matrix-supporting and pore-floating models) show that, our model predicts higher S-wave velocity than matrix-supporting and pore-floating models, but yields similar P-wave velocity estimates as matrix-supporting model. We apply our model to two marine hydrate sites in the Cascadia margin: Site 1245 from Ocean Drilling Program Leg 204 and Site U1328 from International Ocean Drilling Program Expedition 311. Two locations yield similar results: velocity estimates from our model are much closer to downhole measurements than matrix-supporting and pore-floating models. Moreover, we estimate in situ hydrate saturation and clay concentration using our model, matrix-supporting model, and pore-floating model independently, and find that (i) hydrate saturations predicted by our model conform better with the saturations from chloride concentration and (ii) clay contents calculated by our model fit the best with results from smear slide analysis. This study demonstrates that our double-solid-matrix model can be an effective tool to understand the effect of heterogeneously distributed hydrates on velocities, as well as obtain accurate hydrate content in marine isotropic sediments.

SUMMARY: Accurate quantification of natural gas hydrate is essential for resource potential and climate impact assessment. Archie’s empirical equations are commonly used to quantify hydrates from electrical resistivity measurements. One dominant Archie equation parameters, that is, saturation parameter (n), is generally assumed to be constant for different hydrate saturation range for a given reservoir. However, n actually varies with hydrate saturation and morphology, and the exact relationship between n and hydrate saturation or morphology still remains poorly understood, leading to great uncertainties in resistivity-derived saturations. Here we investigate the effect of hydrate saturation and dominant hydrate morphologies on n using well logs from four sites in both fine and coarse-grained sediments: two sites with fluid-displacing hydrate (site W11 from the third Guangzhou Marine Geologic Survey; and Mallik 5L-38 well in the Mackenzie Delta) and two sites with fracture-filling hydrate (site 10 from Indian National Gas Hydrate Program Expedition 01; and site W08 from the second Guangzhou Marine Geologic Survey). We calculated n value using Archie’s law with hydrate saturation determined from velocity. Our results demonstrate a clear negative relationship between hydrate content and n value. Moreover, n estimates from two fracture-filling sites show greater variability compared to fluid-displacing sites. At a fracture-filling hydrate site, site 10, various trends between n and hydrate saturation are possibly caused by the distinct gas compositions of hydrate. Our results demonstrate that significant effects of hydrate morphology and saturation on n that are site specific, and can be used to enhance the accuracy of gas hydrate quantification.

An improved understanding of the effects of gas hydrate presence on seismic attenuation is important for accurate hydrate characterization and quantification. Based on a rock‐physics model recently presented for gas hydrate‐bearing fine‐grained clay‐dominated sediments, here we establish an integrated workflow for surface seismic data from extracting seismic attenuation to estimating gas hydrate concentration (C_h) in the sediment. We apply this workflow to the high‐resolution seismic data acquired at southern Hydrate Ridge, offshore Oregon, to reveal the hydrate distribution and clarify the controlling factor of hydrate formation. We first present an adaptive‐bandwidth spectral ratio method to robustly measure attenuation. The attenuation measurements show that the presence of hydrate suppresses the attenuation of the host sediment. We then calculate C_h by applying the rock‐physics model to the attenuation measurements. The estimated C_h are mostly low (C_h measured from core‐ or well log‐based data. Our result also suggests that the lithology and stratigraphic structures together control the distribution of gas hydrate at Hydrate Ridge, where relatively high C_h is found in the region where a gas‐charged conduit exists and in an anticlinal structure overlying a strong bottom simulating reflection. Adjacent to the anticline, however, a low amount of gas hydrate appears present, possibly due to the gas migration blocked by the anticlinal structure or the lack of gas conduits. Our study offers an effective strategy for detecting and quantifying gas hydrate in fine‐grained clayey sediments through surface seismic data.

By completely solving the second-order full-acoustic wave equation, two-way/full wave equation migration generates images up to arbitrary angles in inhomogeneous media. Full wave equation depth migration (FWDM) necessitates less wavefield storage during imaging, generates clearer imaging results, and has more computational flexibility compared with conventional reverse time migration. Conventionally, FWDM uses Runge-Kutta or two explicit one-way operators to extrapolate the data, which involves either a multirecursive process or complex square root computations of the Helmholtz operator, causing implementation complexities and inconveniences. To overcome these challenges, a novel algorithm called the one-step sparse-matrix method for FWDM (OSM-FWDM) is developed and executed. This method involves a straightforward calculation of the full wave equation propagator through a Taylor expansion and using the so-called precise integration method. The computational efficiency can be significantly improved by using sparse-matrix multiplication in the implementation of wavefield extrapolation. Building upon that, the method serves as a basic solution complemented by FWDM, offering straightforward, efficient, and easy coding characteristics. Our OSM-FWDM method is simpler and faster compared with conventional FWDM, all while maintaining imaging accuracy. In addition, a new solution that uses an extended velocity model to remove source-related artifacts in FWDM is developed and illustrated. These improvements are demonstrated through a series of numerical examples and a real data application.